Process Capability Analysis--Normal and General Non-Normal Distribution (Raw Data) - Options Tab
Select the
Options tab of the
Process Capability Analysis--Normal and General Non-Normal Distribution
Frequency tables and histograms
Use the options in the Frequency tables and histograms group box to control the lower and upper limits, as well as the number of categories to be used in the frequency table or histogram.
You can also select:
- how the observed data are to be categorized (such as how the horizontal x-axis in the histogram is to be scaled)
- where the Sigma limit (process range) indicators are to be placed
- how possibly overlapping text labels are to be placed in the plot
Lower limit
Use the Lower limit box (or the accompanying microscrolls) to enter the desired lower limit for the frequency distribution.
Upper limit
Use the Upper limit box (or the accompanying microscrolls) to enter the desired upper limit for the frequency distribution.
Number of categories
Use the Number of categories box (or the accompanying microscrolls) to enter the desired number of categories for the frequency table and/or histogram.
Show process capability indices
Select the Show process capability indices check box if you want the histograms to include the current process capability indices.
That is, for the normal distribution and when applicable, for the respective non-normal distributions.
Sigma limits at equivalent percentiles
The
Sigma limits at equivalent percentiles check box is only applicable when you are producing a non-normal histogram via the
Advanced, Non-Normal
In that case, the
Sigma limits (which define the process range) can be placed either 1) at the location according to the normal distribution, or 2) at the equivalent percentile value for the respective non-normal distribution. In the first case, the location of the non-normal curve within these limits may provide an indication of how realistic assumptions about the process range are, given the non-normal distribution; in the second case, the indicated (non-normal) process range corresponds to the non-normal process capability indices, which can be displayed via the
Summary: Current variables option on the
Advanced, Non-Normal
"Neat" intervals (appr. No. categs, limits)
Select the Neat intervals (appr. No. categs, limits) check box if you want the range for the respective variable to be divided into approximately the number of requested intervals, subject to the condition that the resulting frequency table or histogram will be based on neat intervals.
Neat in this context means that the category boundaries will be round values ending either in 0, 2, or 5 (e.g., boundaries may be 0.1, 0.2, 0.3, etc.; or 50, 100, 150, etc.). To achieve this, the requested Lower limit, Upper limit, and No. of categories will only be approximated. If the Neat check box is cleared, the boundaries and number of categories will reflect exactly the values you selected, often resulting in "non-neat" category boundaries.
Move titles to avoid overlapping text
The process capability histogram will indicate with lines the nominal, upper, and lower specification limits, as well as the process range (Sigma limits).
When you select the Move titles to avoid overlapping text check box, if any of these lines are very close to each other (which, in practice, should rarely happen), Statistica will move the respective labels for those lines so that they will not overlap. This feature can be turned off by clearing this check box.
Probability and quantile plots
In this group box, select the options for the 2D Probability-Probability Plots and 2D Quantile-Quantile Plots.
Place theoretical quantiles/probs on
Use the options in the Place theoretical quantiles/probs on group box to determine against which axis the theoretical quantiles or probabilities will be plotted.
Horizontal axis
Select the Horizontal axis option button to specify that the theoretical quantiles or probabilities will be plotted against the horizontal axis.
Vertical axis
Select the Vertical axis option button to specify that the theoretical quantiles or probabilities will be plotted against the vertical axis.
Use the options described below to specify the rank and N adjustment factors that are to be used for the computation of quantile plots. These adjustment factors ensure that the p-value for the inverse probability integral for the first (sorted) observation is greater than 0, and for the last (sorted) observation is less than 1 (for additional details, see Chambers, Cleveland, Kleiner, and Tukey, 1983).
Value for adjusting rank
Use the Value for adjusting rank box (and the accompanying microscrolls) to enter the rank adjustment factor you want Statistica to use. The valid range for the rank adjustment factor is .5 <= radj <= 0.
Value for adjusting N
Use the Value for adjusting rank box (and the accompanying microscrolls) to enter the N adjustment factor you want Statistica to use. The valid range for the N adjustment factor 0 < Nadj <= .5.
Average tied ranks
The setting of the Average tied ranks check box determines how tied (equal-value) data points will be treated in P-P and Q-Q plots.
If this check box is cleared, tied data points will be assigned consecutive ranks for the plots and the respective observations will be visible as separate data points in the plots. If this check box is selected, identical average ranks will be assigned to tied observations and the respective points in the plots will overlap (i.e., they will not be identifiable as separate points in the plot).
Reset values
Click the Reset values button to return to the default values for the rank adjustment and N adjustment factors, which are -1/3 = -.3333 and +1/3 = +.3333, respectively.